Method and system for improving accuracy of fatigue driving detection

Abstract

The invention discloses a method and a system for improving the accuracy of a fatigue driving detection. The method comprises: obtaining a face image of a driver photographed by a camera mounted at apredetermined position of the vehicle; judging whether the predetermined position where the camera is mounted is appropriate or not according to the face image, and if appropriate, calculating a current attitude angle of the head of the driver according to the face image; differencing the current attitude angle with a recorded reference attitude angle to obtain an attitude angle difference, and comparing the attitude angle difference with a preset angle determining threshold to determine whether the driver is currently in a fatigue driving state or not according to the comparison result. According to the technical scheme of the method and the system for improving the accuracy a fatigue driving detection, the mounting position of the camera which is taken as the foundation of a fatigue driving detection is firstly calibrated before the fatigue driving detection is performed, that is, whether the mounting position of the camera is appropriate or not is determined, so that the problems ofthe misdetection and leakage detection of fatigue driving due to an inappropriate mounting position of the camera are avoided, and the accuracy of the fatigue driving detection is ensured and improved from the source.

Description

technical field [0001] The invention relates to the technical field of safe driving detection, in particular to a method and system for improving the accuracy of fatigue driving detection. Background technique [0002] With the development and application of machine learning and computer vision technology in human behavior detection, algorithms can now capture images of drivers closing their eyes, yawning, bowing their heads, and looking left and right through the camera, and then analyze these abnormal states of the driver. Driving alarm. The premise of judging whether the driver is looking down or looking left and right is that the installation position of the camera is fixed with the driver, so as to ensure accuracy. For example, the camera is uniformly installed in front of or on the right side of the driver. If the camera is installed at any position relative to the driver, since the attitude measurement is also relative, the probability of false detection and missed d...

AI Technical Summary

Problems solved by technology

If the camera is installed at any position relative to the driver, the measurement of attitude is also relative, which greatly increases the probability of false detection and missed detection

Images